20VC Redpoint's Tom Tunguz on Winning with Data How To Gain A Competitive Advantage & Dominate Markets with Data and 5 Steps To Create A Data Driven Company

In this episode of "20 minutes VC," host Harry Stebbings interviews Tom Tunguz, a partner at Redpoint Ventures and author of "Winning with Data." They discuss the transformative power of data in business, with Tunguz sharing his insights on how data-driven companies iterate faster and make more informed decisions. They explore the importance of structured interviews in hiring, the challenges of distinguishing correlation from causation in data analysis, and the necessity of emotions in decision-making. Tunguz also highlights Thredup as a prime example of a data-driven company and introduces Dremio, his latest investment focused on data middleware. The conversation underscores the evolving role of data in shaping company culture, strategy, and operational efficiency.

• Harry Stebings introduces the feature week with Redpoint, a leading VC fund in Silicon Valley.
• Tom Tunguz, a partner at Redpoint and a prominent thinker in SaaS, is the guest on the show.
• Tunguz has recently released a book titled "Winning with Data," which captivated Harry despite his usual disinterest in books.
• The episode will explore key aspects of Tunguz's book and his insights on data.

"And you can add me at h Stebings on Snapchat. And for the show this week we have a very special feature week as we feature one of the valley's leading vc funds, Redpoint. And joining me today from Redpoint, we have none other than Tom Tunguz, partner at Redpoint."

The quote introduces the episode's focus on Redpoint and its partner, Tom Tunguz, highlighting his expertise in SaaS and his new book.

• Cooley, a global law firm specializing in startups and venture capital, is highlighted for its experience with VCs.
• Cooley has a significant history in forming venture capital funds and supporting companies through their lifecycle.
• Eve, a UK-based direct consumer mattress company, offers a unique trial period and savings to customers, emphasizing the importance of operational support and rest for productivity.

"Cooley are the global law firm built around startups and venture capital. Since forming the first venture fund in Silicon Valley, Cooley has formed more venture capital funds than any other law firm in the world."

The quote explains Cooley's role and expertise in the venture capital space, indicating its importance to startups and VC operations.

• Tunguz's first "aha moment" with data came at age 17 while working for a law firm in Santiago, Chile.
• He founded a company with his father to address the law firm's billing delays.
• Tunguz demonstrated the ability to identify top-performing attorneys using billing data, impressing the managing partner.
• At Google, Tunguz worked on the Adsense team and used data to compete with Yahoo by tracking ad placements.

"With just a little bit of data, the stuff that we were collecting about who was billing hours and the interactions between the attorneys and the clients, we knew a whole lot."

The quote exemplifies the early recognition Tunguz had of the power of data in identifying performance and making informed decisions.

• Tunguz has been passionate about data for a long time, observing the advantages of data-driven companies.
• Data-driven companies iterate faster and have a deeper understanding of their operations.
• Tunguz's experience with Looker, a fast-growing BI company, provided many stories of businesses transforming through data.
• The book was written to share these stories and the impact of data on various industries.

"The biggest difference among them is that data-driven companies tend to iterate faster."

This quote emphasizes the competitive edge that data-driven companies have due to their ability to quickly adapt and improve.

• Best-in-class data-driven companies share three common practices:
  1. They establish dedicated data teams.
  2. They democratize data access across the organization.
  3. They cultivate a culture that values data-driven decision-making.

"Yeah, there are three things they do. The first is they have a data team, and there's this data..."

The quote introduces the three key practices that successful data-driven companies implement to leverage data effectively.

• Data teams are composed of experts who know where data is stored and how to analyze it.
• Their primary function is to educate others in the organization through office hours, tech talks, and similar activities.

"They're the experts. They know where the data is stored, they understand how to perform analysis, but their main function is really education."

This quote highlights the role of data teams as educators within a company, emphasizing their expertise in data storage and analysis.

• The concept of a data dictionary was pioneered by the head data scientist at Warby Parker to address the issue of inconsistent metric definitions across different teams.
• A data dictionary serves as a central reference that defines all key metrics for the business, preventing communication breakdowns due to differing definitions.

"So the second common characteristic is there's a single person or team that creates this data dictionary that defines all the key metrics for the business once and for all."

The quote explains the importance of having a unified set of definitions for key metrics within a business to avoid miscommunication and inconsistencies.

• Data pipelines are crucial for transferring data from the depths of the business to the front lines.
• Many current data pipelines are struggling to handle the volume of data they are required to process.
• The concept of a "data breadline" illustrates the invisible line where employees wait to get the data they need, often requiring them to curry favor with data teams.

"And the third thing is what we call in the book a data pipeline, which is a way of getting the data from the deep recesses of the business out to the front lines."

This quote introduces the data pipeline as a means of moving data within a company, which is essential for making data accessible to those who need it.

• Data-driven companies often have a culture of questioning and transparency around data.
• It's challenging to maintain transparency without causing internal issues, but it begins with a questioning culture and leadership that prioritizes data over opinions.

"There's this famous quote that said, it basically said, if we have data, let's go with the data, but if we have opinions, let's go with mine."

The quote underscores the prioritization of data over personal opinions in data-driven companies, highlighting the cultural value placed on evidence-based decision-making.

• Changing a company's culture to be more data-driven starts with leadership asking open-ended questions.
• Leaders should empower their teams to seek data-driven answers, creating a bottom-up and top-down infusion of data into the company's culture.

"I think the first thing is the leadership decides to start asking questions, and they ask open-ended questions."

This quote emphasizes the role of leadership in fostering a data-driven culture by encouraging curiosity and the pursuit of data-supported answers.

• The idea of "not listening to your boss" is about using data to form one's own conclusions and having a strong point of view supported by data.
• Authoritative figures are important, but employees should be encouraged to think critically and independently.

"But the important thing about not listening to your boss is really coming to your own conclusions, using data to support your conclusions, and then having a point of view."

The quote captures the essence of a data-driven approach where employees are encouraged to rely on data rather than solely on directives from superiors.

• Structured interviews and standardized tests can significantly improve the predictive capability of hiring success for data-driven roles.
• Defining the role and success criteria, followed by structured questions, can enhance hiring efficiency.
• Metrics such as the recruiting sales cycle are crucial for an effective recruiting process.

"Everybody asks the same set of questions. And those questions are based upon the characteristics that are going to define success in this role."

This quote explains the methodology behind structured interviews, where consistency in questioning aligns with the specific success factors of the role.

• There is an implicit question about the extent to which human emotion should influence data-driven decisions.
• The balance between data and human judgment is a consideration in the data-driven approach.

"And it's to what extent do we actually allow human emotion in to drive our decisions, too, pa[rtially cut off]"

This incomplete quote introduces a contemplation on the role of human emotion in the context of data-driven decision-making.

• Emotions play a crucial role in decision-making processes.
• The inability to make decisions can be linked to the absence of emotions, as illustrated by the case of a man in Germany.
• Even with structured interviews and scoring systems, emotional components and advocacy within the hiring committee are important.
• The story of the German man and Google's hiring process both highlight the necessity of emotions in decision-making.

So there was this guy in Germany who was struck by lightning, and the part of his brain that created emotions was destroyed... He actually couldn't make a single decision. And what they realized is that we never have perfect information. And the way that we get from not making a decision to making a decision is we use our emotions.

This quote emphasizes that emotions are integral to decision-making because they help bridge the gap when information is incomplete.

And I think that's when you're not hiring exclusively based upon the metrics that are coming out of the interviews. There needs to be an advocate inside of the hiring committee, and people need to argue back and forth. And I think that's where the emotional component definitely comes in.

The quote suggests that hiring decisions should not solely rely on interview metrics but also on discussions and advocacy within the hiring committee, where emotions and interpersonal dynamics play a role.

• Data analysis requires discipline and an understanding of complex concepts like correlation versus causation.
• Experts can be wrong, as shown by the Monty Hall problem and the backlash faced by the journalist who correctly solved it.
• Two common biases in data analysis are anchoring bias and availability bias.
• The illusion of validity is a cognitive bias where people believe they can predict outcomes when they cannot, as demonstrated by Daniel Kahneman's experience at West Point.

Getting to the point of the question, the most important one, and I think that the skill that most often comes up in data analysis is correlation versus causation.

Tom Tunguz highlights the importance of distinguishing between correlation and causation, which is a fundamental skill in data analysis.

And then the second one is this notion of anchoring bias... Telling you that number and putting a number in your mind ahead of time, you are already thinking around that number, and that kind of biases you.

The quote explains the concept of anchoring bias, where initial information provided influences subsequent judgments and decisions.

Yeah, the availability bias is basically, this is founded by Daniel Kahneman. It's basically, if it's something that's easily remembered, then you perceive it to be more likely.

Tom Tunguz describes availability bias, where people overestimate the likelihood of events that are more memorable or easily recalled.

And so the illusion of validity in this case is that Kahneman and his assistant thought they could predict when in fact, they couldn't. Right? They had no idea.

This quote illustrates the illusion of validity, where individuals have an unfounded confidence in their predictive abilities.

• Testing assumptions through small-scale experiments is a method to prevent the illusion of validity.
• Google's approach to validating assumptions can be applied to prevent a false sense of knowledge in decision-makers.

Yeah, I think that what you do is you do what Google does or did when I was there, which is you create a hypothesis and you say, let's run a small scale experiment just to make sure that it's true. Basically what you want to do is you want to test your assumptions.

Tom Tunguz suggests that to avoid the illusion of validity, one should test assumptions with experiments, similar to practices at Google.

• Writing a book is a significantly different process compared to writing a blog post.
• The challenge of writing a book involves dealing with a much larger word count and a more extensive writing project.

No verbatim quote provided for this theme.

• The challenge of creating content that is deep enough to engage but not too deep to bore the reader.
• The absence of instant feedback, unlike a blog, which makes it difficult to gauge reader interest and make corrections.

"The biggest challenge was trying to figure out how do you go deep enough into a topic to be interesting to the reader, but not so deep that you're bored and you don't have instant feedback like a blog, and you can't correct it, so you just don't know."

This quote emphasizes the difficulty in finding the right balance in depth of content to maintain reader interest without the benefit of immediate feedback.

• Tom Tunguz had a 90-day deadline to write his book.
• The writing was done from November 1 to February 1, during nights and weekends.
• Tom's wife supported him by taking care of their children during this time.

"I had 90 days to write it. You said? Yeah."

Tom confirms the tight timeframe he had to complete his book.

"Yeah, we wrote it from November 1 to February 1."

This quote specifies the exact period Tom spent writing the book.

"Nights and weekends. Yeah. My lovely wife watched our boys."

Tom acknowledges the personal time commitment and family support required in the book writing process.

• Tom used Scrivener and Hemingway as tools for writing his book.
• Scrivener helped in organizing content like individual blog posts and compiling them into a structure.
• Hemingway was used for spelling and punctuation.

"You know, there's this great software called Scrivener."

Tom mentions Scrivener as a key tool in his writing process.

"Hemingway is great, too."

Tom also acknowledges the usefulness of Hemingway in his writing process.

• The realization that no one has complete knowledge on how to do everything correctly and that methodologies are always evolving.
• The importance of understanding data and statistics in decision-making, illustrated by a historical example from World War II.
• Business intelligence and data analysis are relatively young fields with much room for growth and learning.

"My biggest takeaway is nobody knows how to do all this stuff right. And it's always changing."

Tom reflects on the continual learning and adaptation required in fields like data analysis and business intelligence.

"There's one story about in the second world war...So the planes that were coming back had holes in the fuselage. And so you don't need to armor the fuselage. You need to armor the wings or whatever it was."

Tom uses an anecdote from World War II to explain a fundamental lesson in statistics and its application to business intelligence.

• Tom is impressed with Thredup for its data-driven operations.
• Thredup processes a large volume of clothing daily, and its operations are measured meticulously.
• The company's data-driven approach has a significant impact on its financial margins.

"Everything at Thredup is measured."

Tom highlights Thredup's comprehensive use of data in its operations.

• Human resources software companies have the potential to innovate with data.
• Traditionally, marketing and HR heads lacked data at board meetings, but marketing has since become fully instrumented.
• The next opportunity lies in equipping HR heads with data.

"The next great opportunity is helping the head of people with data."

Tom identifies human resources as the next frontier for data-driven decision-making within companies.

• Tom's latest public investment is in Dremio, a data middleware company.
• Dremio enables seamless data movement and helps with data discovery within companies.
• The company addresses the challenge of connecting and finding relevant data sets for analysis.

"They basically allow you to move data in really beautiful and seamless ways, and they allow people inside of companies to find the data they need access to."

Tom explains the value proposition of Dremio and why he chose to invest in it.

• Tom expresses gratitude for Harry's appreciation of his book.
• Harry's mention of reading Tom's book from start to finish is significant as it was the first book he completed in about two years.

"I'm honored. Thank you so much, Harry."

Tom shows appreciation for the positive feedback on his book from Harry.

• Harry promotes the Twenty Minute VC platform and encourages listeners to purchase Tom's book.
• Harry highlights sponsors such as Cooley, a law firm specializing in startups and venture capital, and Eve, a mattress company.
• Upcoming episodes and guests are teased, including Ryan Sarver from Redpoint.

"And if you love the episode, then head over to the twentyminutevc.com where you can find the links to purchase the book."

Harry directs listeners to the podcast website for further engagement and to access the book.

"And do not forget what I said earlier. Funds would be nothing without operational support behind them."

Harry discusses the importance of operational support for venture capital funds, highlighting Cooley's services.

"Eve is the UK's number one direct consumer mattress company."

Harry promotes Eve as a sponsor and emphasizes their customer trial and return policy.

"I'm also delighted to announce that we will have Ryan Sarver on the show."

Harry teases an upcoming episode, creating anticipation for future content.